Pixel-based Classification Methods Research Articles

Winter Wheat Mapping in Shandong Province of China with Multi-Temporal Sentinel-2 Images

Wheat plays an important role in China’s and the world’s food supply, and it is closely related to economy, culture and life. The spatial distribution of wheat is of great significance to the rational planning of wheat cultivation areas and the improvement of wheat yield and quality. The current rapid development of remote sensing technology has greatly improved the efficiency of traditional agricultural surveys. The extraction of crop planting structure based on remote sensing images and technology is a popular topic in many researches. In response to the shortcomings of traditional methods, this research proposed a method based on the fusion of the pixel-based and object-oriented methods to map the spatial distribution of winter wheat. This method was experimented and achieved good results within Shandong Province. The resulting spatial distribution map of winter wheat has an overall accuracy of 92.2% with a kappa coefficient of 0.84. The comparison with the actual situation shows that the accuracy of the actual recognition of winter wheat is higher and better than the traditional pixel-based classification method. On this basis, the spatial pattern of winter wheat in Shandong was analyzed, and it was found that the topographic undulations had a great influence on the spatial distribution of wheat. This study vividly demonstrates the advantages and possibilities of combining pixel-based and object-oriented approaches through experiments, and also provides a reference for the next related research. Moreover, the winter wheat map of Shandong produced in this research is important for yield assessment, crop planting structure adjustment and the rational use of land resources.

Aboveground Biomass Inversion Based on Object-Oriented Classification and Pearson–mRMR–Machine Learning Model

Cities play a crucial role in the carbon cycle. Measuring urban aboveground biomass (AGB) is essential for evaluating carbon sequestration. Satellite remote sensing enables large-scale AGB inversion. However, the apparent differences between forest and grassland biomass pose a significant challenge to the accurate estimation of urban AGB using satellite-based data. To address this limitation, this study proposed a novel AGB estimation method using the eastern part of the Zhahe mining area in Huaibei City as the study area, which integrates land cover classification, feature selection, and machine learning modelling to generate high quality biomass maps of different vegetation types in an urban area with complex feature distribution. Utilizing the GEE platform and Sentinel-2 image, we developed an object-oriented machine learning classification algorithm, combining SNIC and GLCM to extract vegetation information. Optimal feature variables for forest and crop-grass AGB inversion were selected using the Pearson–mRMR algorithm. Finally, we constructed nine machine learning models for AGB inversion and selected the model with the highest accuracy to generate the AGB map of the study area. The results of the study are as follows: (1) Compared with the pixel-based classification method, the object-oriented classification method can extract the boundaries of different vegetation types more accurately. (2) Forest AGB is strongly correlated with vegetation indices and physiological parameters, while agri-grass AGB is primarily associated with vegetation indices and vegetation physiological parameters. (3) For forest AGB modelling, the RF-R model outperforms other machine learning models with an R2 of 0.77. For agri-grass AGB modelling, the XGBoost-R model is more accurate, with an R2 of 0.86. (4) The mean forest AGB in the study area was 4.60 kg/m2, while the mean agri-grass AGB was 0.71 kg/m2. High AGB values were predominantly observed in forested areas, which were mainly distributed along roads, waterways, and mountain ranges. Overall, this study contributes to a better understanding of the health of local urban ecosystems and provides valuable insights for ecosystem protection and the sustainable use of natural resources.

Pixel-based classification method for earthquake-induced landslide mapping using remotely sensed imagery, geospatial data and temporal change information

Pixel-based classification method for earthquake-induced landslide mapping using remotely sensed imagery, geospatial data and temporal change information

Decision Fusion at Pixel Level of Multi-Band Data for Land Cover Classification-A Review.

According to existing signatures for various kinds of land cover coming from different spectral bands, i.e., optical, thermal infrared and PolSAR, it is possible to infer about the land cover type having a single decision from each of the spectral bands. Fusing these decisions, it is possible to radically improve the reliability of the decision regarding each pixel, taking into consideration the correlation of the individual decisions of the specific pixel as well as additional information transferred from the pixels' neighborhood. Different remotely sensed data contribute their own information regarding the characteristics of the materials lying in each separate pixel. Hyperspectral and multispectral images give analytic information regarding the reflectance of each pixel in a very detailed manner. Thermal infrared images give valuable information regarding the temperature of the surface covered by each pixel, which is very important for recording thermal locations in urban regions. Finally, SAR data provide structural and electrical characteristics of each pixel. Combining information from some of these sources further improves the capability for reliable categorization of each pixel. The necessary mathematical background regarding pixel-based classification and decision fusion methods is analytically presented.

Sentinel-2 image based smallholder crops classification and accuracy assessment by UAV data

Timely and accurate extraction of crop planting units plays a critical role in crop yield estimation, soil management, food supplies, and disaster warnings. However, precisely mapping crop types is challenging in smallholder farming systems due to heterogeneous and mixed pixels, where field sizes are small, and crop types are very diverse. In this paper, the crop type system of Huocheng County in northwest China as an example area, crop classification feature variables are constructed using Sentinel-2 remote sensing images, and combining the ReliefF algorithm, three feature selection classification models are established by 3092 crop type sampling points data. The crop planting units are extracted using pixel-based and object-based classification methods, respectively, and unmanned aerial vehicle (UAV) data assess the accuracy as ground truth. The Sentinel-2 image based smallholder crops classification results indicate that: The effective combination of optimal input feature variables selection and classification models significantly improves crop classification accuracy. This is particularly evident when integrating this approach into pixel-based classification, addressing challenges such as low pixel classification accuracy in regions of this type. The validation results from UAV data also confirm these findings. At validation points 2, 3, and 4, pixel-based classification demonstrates significantly higher accuracy compared to object-based classification, with pixel-based classification accuracy improving by 9.16%, 7.83%, and 31.83%, respectively. Although the UAV validation accuracy does not reach a high level, pixel-based classification remains the optimal choice for smallholder crop classification. This method offers new insights and references for research related to the classification of smallholder crops with complex planting structures.

Comparison of Automatic Classification Methods for Identification of Ice Surfaces from Unmanned-Aerial-Vehicle-Borne RGB Imagery

This article describes a comparison of the pixel-based classification methods used to distinguish ice from other land cover types. The article focuses on processing RGB imagery, as these are very easy to obtained. The imagery was taken using UAVs and has a very high spatial resolution. Classical classification methods (ISODATA and Maximum Likelihood) and more modern approaches (support vector machines, random forests, deep learning) have been compared for image data classifications. Input datasets were created from two distinct areas: The Pond Skříň and the Baroch Nature Reserve. The images were classified into two classes: ice and all other land cover types. The accuracy of each classification was verified using a Cohen’s Kappa coefficient, with reference values obtained via manual surface identification. Deep learning and Maximum Likelihood were the best classifiers, with a classification accuracy of over 92% in the first area of interest. On average, the support vector machine was the best classifier for both areas of interest. A comparison of the selected methods, which were applied to highly detailed RGB images obtained with UAVs, demonstrates the potential of their utilization compared to imagery obtained using satellites or aerial technologies for remote sensing.

An advanced coverage estimation method to quantify biological soil crust coverage using Sentinel-2 imagery in desert and sandy land of China

ABSTRACT Monitoring the distribution and area change of biological soil crusts (BSCs) can enhance our understanding of the interactions between nonvascular plants and the environment in drylands. However, using only pixel-based binary classification methods results in large-area estimation errors at large scales. The lack of available calculation methods for directly measuring BSC coverage using multispectral satellite images makes it challenging to obtain BSC area data for further studies at large scales. To address these issues, this study developed feature space conceptual models for desert and sandy land based on the characteristics of BSC in drylands. The desert feature space comprised the normalized difference vegetation index (NDVI) combined with the brightness index (BI), encompassing moss, lichen, and non-BSC. The sandy land feature space relied on the biological soil crust index (BSCI) and the NDVI, including vegetation, mixed BSCs and sandy soil. Using Sentinel-2 satellite imagery and a spectral unmixing model, the abundance of BSCs was quantified in four BSC growth areas located in the Gurbantunggut Desert and Mu Us Sandy Land of China. Validation of the method indicated that the root mean square error (RMSE) of the BSC coverage estimation results was 10% and 8% in desert and sandy land, respectively (estimation accuracies of 79% and 81%, respectively). This demonstrated that the proposed method can effectively estimate BSC coverage at a subpixel scale. The resulting BSC coverage data can provide the possibility to evaluate the functions of regional ecosystems.

Research on the identification of land types and tree species in the Engebei ecological demonstration area based on GF-1 remote sensing

Identifying forest types is crucial in satellite remote sensing monitoring. Research focusing on the identification of forest tree species using high-resolution remote sensing images is on the rise. Traditional pixel-based classification methods have not been successful in fully utilizing the rich spatial and texture feature information present in the image. Further, these methods are plagued by “classification noise”, which impedes the precise extraction of information regarding forest tree species. The object-oriented classification method, on the other hand, offers an extraction methodology rooted in image segmentation object features. It efficiently leverages the spectral, texture, and spatial geometry information of high-resolution remote sensing images. In this study, an initial application of the object-oriented, multi-level information extraction was conducted on ground objects within the Engebei ecological demonstration area. This was performed utilizing Gaofen (GF)-1 remote sensing images. Following this, a multi-scale segmentation algorithm was utilized to establish different segmentation scales relative to the features of the different objects. The optimal segmentation scale, shape factor, and compactness factor were determined to be 222/0.5/0.5, 167/0.5/0.6, and 81/0.4/0.6, respectively, for the three levels. Subsequently, training samples were used to select optimal features during the classification process. The nearest neighbor classification and random forest classification were performed respectively. The results indicate an overall classification accuracy of 88.58% and 87.26% for the nearest neighbor classification of remote sensing images in 2020 and 2021 respectively. Meanwhile, the overall classification accuracy for the random forest classification was 92.95% and 92.02% respectively. The Kappa coefficients for the nearest neighbor classification were 0.86 and 0.85, whereas those for the random forest classification were 0.92 and 0.90. These outcomes underscore the utility of the object-oriented classification method in enhancing the accuracy of forest type identification.

Spatial distribution of benthic habitats in Kapota Atoll (Wakatobi National Park, Indonesia) using remote sensing imagery

Abstract. Hamuna B, Pujiyati S, Gaol JL, Hestirianoto T. 2023. Spatial distribution of benthic habitats in Kapota Atoll (Wakatobi National Park, Indonesia) using remote sensing imagery. Biodiversitas 24: 3700-3707. This study aims to classify and map benthic habitats in Kapota Atoll, Wakatobi National Park, using Sentinel-2A satellite imagery, that has undergone both atmospheric and geometric corrections. Sentinel-2A satellite imagery was processed using pixel-based classification method. This study used two machine learning algorithms: support vector machine (SVM) and random forest (RF). This study used nine classified benthic habitat classes, namely five homogeneous benthic classes (sand, live coral, dead coral, rubble, and dense seagrass) and four mixed benthic classes (mixed rubble and sand, mixed sand and rare seagrass, mixed medium seagrass and sand, and a mixture of rubble, dead coral, and seagrass). The RF algorithm produced benthic habitat maps with a higher overall accuracy than the SVM algorithm, which was 65.31% and 63.51%, respectively. These accuracy values have met the standards of accuracy for mapping benthic habitats that apply in Indonesia. The SVM algorithm classification results showed that the benthic habitat of the mixed class of medium seagrass and sand was more dominant than other benthic habitat classes. In contrast, the RF algorithm showed the class dense seagrass was more dominant. The selection of the appropriate algorithm for processing satellite imagery was found to be a significant factor in producing an accurate benthic habitat map.

Integrated Method for Classifying Medium Resolution Satellite Remotely Sensed Imagery into Land Use Map

There are several remotely sensed images of varied resolutions available. As a result, several classification techniques exist, which are roughly classified as pixel-based and object-based classification methods. Based on the foregoing, this study provided an integrated method of deriving land use from a coarse satellite image. Location coordinates of the land uses were acquired with a handheld Global Positioning System (GPS) instrument as primary data. The study classified the image quantitatively (pixel-based) into built-up, water, riparian, cultivated, and uncultivated land cover classes with no mixed pixels, and then qualitatively into educational, commercial, health, residential, and security land use classes that were conflicting due to spectral similarity. The total accuracy and kappa coefficient of the pixel-based land cover classification were 92.5% and 94% respectively. The results showed that residential land use occupied an area of 5500.01ha, followed by educational (2800.69ha); security (411.27ha); health (133.88ha); and commercial (109.01ha) respectively. The integrated method produces a crisp-appearance like the object-based image classification method. It eliminates the "salt and pepper" appearance that a traditional pixel-based classification would have. The output can be a vector or raster model depending on the purpose for which it is created.

Deep Learning of High-Resolution Unmanned Aerial Vehicle Imagery for Classifying Halophyte Species: A Comparative Study for Small Patches and Mixed Vegetation

Recent advances in deep learning (DL) and unmanned aerial vehicle (UAV) technologies have made it possible to monitor salt marshes more efficiently and precisely. However, studies have rarely compared the classification performance of DL with the pixel-based method for coastal wetland monitoring using UAV data. In particular, many studies have been conducted at the landscape level; however, little is known about the performance of species discrimination in very small patches and in mixed vegetation. We constructed a dataset based on UAV-RGB data and compared the performance of pixel-based and DL methods for five scenarios (combinations of annotation type and patch size) in the classification of salt marsh vegetation. Maximum likelihood, a pixel-based classification method, showed the lowest overall accuracy of 73%, whereas the U-Net classification method achieved over 90% accuracy in all classification scenarios. As expected, in a comparison of pixel-based and DL methods, the DL approach achieved the most accurate classification results. Unexpectedly, there was no significant difference in overall accuracy between the two annotation types and labeling data sizes in this study. However, when comparing the classification results in detail, we confirmed that polygon-type annotation was more effective for mixed-vegetation classification than the bounding-box type. Moreover, the smaller size of labeling data was more effective for detecting small vegetation patches. Our results suggest that a combination of UAV-RGB data and DL can facilitate the accurate mapping of coastal salt marsh vegetation at the local scale.

Endmember variability based abundance estimation of red and black soil over sparsely vegetated area using AVIRIS-NG hyperspectral image

Visible-near-and-short-wave-infrared hyperspectral images (HSI) have proven helpful for mapping the soil types over bare soils pixels. However, the accuracy of the traditional pixel-based classification methods decreases due to the spectral mixing between the significant agricultural features such as vegetation, soil and crop residue. In this context, spectral unmixing algorithms are handy for estimating the sub-pixel abundances of soil over sparsely vegetated areas. Most of the spectral unmixing methods focus on analyzing the HSI by considering the pixels as independent entities. However, in the HSIs of agricultural fields, the endmembers are spatially and temporally variable. It is also known that there exists a strong spatial correlation among neighbourhood pixels over agricultural fields. Therefore, both endmember variability and the neighbourhood spatial-spectral information are critical to estimate soil abundance accurately. Here we address the issue of estimating the abundances of two major soil types with spectral variability. The proposed approach combines the endmember bundle extraction with the spectral-spatial weighted unmixing approach (SSWU-SV). Experimental analysis has been carried out on the airborne HSI acquired by airborne-visible-and-infrared-imaging- spectrometer-next-generation (AVIRIS-NG) sensor. The quantitative analysis reveals that the proposed method consistently achieves a better unmixing performance than the traditional linear mixing model in terms of spectral angle mapper (SAM), and root-mean-square error (RMSE). Our results also indicate the saturation of normalized difference vegetation index (NDVI) at high vegetative fractions obtained from SSWU-SV.

Mapping global distribution of mangrove forests at 10-m resolution

Mangrove forests deliver incredible ecosystem goods and services and are enormously relevant to sustainable living. An accurate assessment of the global status of mangrove forests warrants the necessity of datasets with sufficient information on spatial distributions and patch patterns. However, existing datasets were mostly derived from ∼30 m resolution satellite imagery and used pixel-based image classification methods, which lacked spatial details and reasonable geo-information. Here, based on Sentinel-2 imagery, we created a global mangrove forest dataset at 10-m resolution, namely, High-resolution Global Mangrove Forests (HGMF_2020), using object-based image analysis and random forest classification. We then analyzed the status of global mangrove forests from the perspectives of conservation, threats, and resistance to ocean disasters. We concluded the following: (1) globally, there were 145,068 km2 mangrove forests in 2020, among which Asia contained the largest coverage (39.2%); at the country level, Indonesia had the largest amount of mangrove forests, followed by Brazil and Australia. (2) Mangrove forests in South Asia were estimated to be in the better status due to the higher proportion of conservation and larger individual patch size; in contrast, mangrove forests in East and Southeast Asia were facing intensive threats. (3) Nearly, 99% of mangrove forest areas had a patch width greater than 100 m, suggesting that nearly all mangrove forests were efficient in reducing coastal wave energy and impacts. This study reports an innovative and up-to-date dataset and comprehensive information on mangrove forests status to contribute to related research and policy implementation, especially for supporting sustainable development.

Convolutional Neural Network Shows Greater Spatial and Temporal Stability in Multi-Annual Land Cover Mapping Than Pixel-Based Methods

Satellite imagery is the only feasible approach to annual monitoring and reporting on land cover change. Unfortunately, conventional pixel-based classification methods based on spectral response only (e.g., using random forests algorithms) have shown a lack of spatial and temporal stability due, for instance, to variability between individual pixels and changes in vegetation condition, respectively. Machine learning methods that consider spatial patterns in addition to reflectance can address some of these issues. In this study, a convolutional neural network (CNN) model, U-Net, was trained for a 500 km × 500 km region in southeast Australia using annual Landsat geomedian data for the relatively dry and wet years of 2018 and 2020, respectively. The label data for model training was an eight-class classification inferred from a static land-use map, enhanced using forest-extent mapping. Here, we wished to analyse the benefits of CNN-based land cover mapping and reporting over 34 years (1987–2020). We used the trained model to generate annual land cover maps for a 100 km × 100 km tile near the Australian Capital Territory. We developed innovative diagnostic methods to assess spatial and temporal stability, analysed how the CNN method differs from pixel-based mapping and compared it with two reference land cover products available for some years. Our U-Net CNN results showed better spatial and temporal stability with, respectively, overall accuracy of 89% verses 82% for reference pixel-based mapping, and 76% of pixels unchanged over 33 years. This gave a clearer insight into where and when land cover change occurred compared to reference mapping, where only 30% of pixels were conserved. Remaining issues include edge effects associated with the CNN method and a limited ability to distinguish some land cover types (e.g., broadacre crops vs. pasture). We conclude that the CNN model was better for understanding broad-scale land cover change, use in environmental accounting and natural resource management, whereas pixel-based approaches sometimes more accurately represented small-scale changes in land cover.

Optimal Scale Selection and an Object-Oriented Method Used for Measuring and Monitoring the Extent of Land Desertification

Desertification has become a major problem in the field, affecting both the global ecological environment and economy. The effective monitoring of desertified land is an important prerequisite for land desertification protection and governance. With the aim of addressing the problems of spectral confusion as well as the salt and pepper phenomenon concerning the successful extraction of desertification information by utilizing the pixel-based method in the studies, Landsat remote sensing images obtained from the year 2001 to 2021 were selected in this study as the data source, and then, the object-oriented random forest classification method was improved by using different optimal segmentation scale selection techniques and combining multi-thematic index characteristics for measuring the extent of land desertification. Finally, the improved method was applied to study the dynamic changes in desertification in the Mu Us Sandy Land Ecological Function Reserve. The results show that the optimal scale determined by different optimal segmentation scale selection methods is not entirely consistent, and a minor scale should be selected as the optimal scale. Compared with the pixel-based classification method, the overall accuracy of object-oriented classification based on the optimal segmentation scale was improved by 8.06%, the Kappa coefficient increased by 0.1114, and the salt and pepper phenomenon was significantly reduced. From 2001 to 2021, the area of desertified land decreased by 587.12 km2 and the area of severely desertified land decreased by 4115.92 km2, indicating that the control effect was remarkable. This study can provide effective decision-making evidence and support for the successful governance of desertification.

Multi-Sensor Image Feature Fusion via Subspace-Based Approach Using $\ell _{1}$-Gradient Regularization

Image fusion is a technique of combining two or more images into a single image which is more informative from an interpretation point of view. With the rapid development of different synthetic aperture radar sensing satellites capturing information from the earth by measuring energy in different portions of the electromagnetic spectrum (narrow/wide-band), complementary information about the area captured by different satellites is available (e.g. high-resolution spectral and RGB images). However, the estimation of the full-resolution image may not be necessary for inference approaches, including the pixel-based classification. Instead, it is desirable to extract the relevant information embedded in the available data to improve the inference capabilities. This work proposes a computational framework to estimate features with high-spatial-resolution and appropriate spectral content by combining information from a multi-sensor system. The considered multi-sensor setup is a hyperspectral imaging system with a complementary RGB sensor. The proposed framework first extracts spatial features from the RGB image using morphological profiles. Then, the fusion model assumes that the extracted features, and the hyperspectral measurements, lie in different subspaces matrices. In addition, this work developed a joint optimization scheme to solve the feature fusion problem by integrating the alternating direction method of multipliers with the block coordinate descent method. The alternating optimization method estimates the spatial features in the fusion model by penalizing the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\ell _{1}$</tex-math></inline-formula> -norm of the spatial gradient magnitudes. The quality of extracted features is measured in terms of supervised pixel-based classification methods. Extensive simulations show that the proposed approach outperforms other state-of-the-art methods in terms of classification accuracy.

Feasibility assessment of tree-level flower intensity quantification from UAV RGB imagery: A triennial study in an apple orchard

A timely and accurate spatial inventory of flowering characteristics benefits both the floral phenology monitoring in ecology and various crop management activities in agricultural systems. Recent advancement has proven the superiority of computer vision in flower classification at image level. Yet progress in the flowering intensity estimation at tree level is much less and still far from satisfactory. To tackle this problem, a novel approach was designed for the use of single raw aerial images to quantify flower intensity. With pre-prepared dataset, flower-associated pixels were extracted for individual trees using a pixel-based classification method, the color thresholding. Next, three flowering indices retrieved from unmanned aerial vehicle (UAV) were evaluated, the index percentage (IPG), index pixel (IP), and index area (IA). Finally, linear correlation of the flowering indices to flower cluster number and expert-assessed floridity recorded in the field were calculated. Results indicated that IPG yielded the highest correlation to flower cluster (R2 = 0.93, RMSE = 8) and floridity estimation (R2 = 0.78, RMSE = 0.9). A UAV-based floridity scoring method was also designed for automatic estimation tasks in practice, and a comparable and even better performance to the expert-based approach was demonstrated. Furthermore, effects of vertical (nadir) and horizontal (angular) overlapping of flower clusters within the canopy were evaluated, showing excellent potential to improve the estimation accuracy.

Superpixel segmentation integrated feature subset selection for wetland classification over Yellow River Delta.

Wetlands are one of the world's most significant and vulnerable ecosystems. The wetlands of the Yellow River Delta are subject to multiple influences of ocean tidal action and the massive sediment deposits of the Yellow River, resulting in a more complex and unstable composition of land cover types. To better distinguish the wetlands in the region, we conducted the classification using an object-oriented combined with feature preference machine learning approach. To alleviate the pretzel phenomenon in pixel-based classification, a superpixel segmentation method using the watershed algorithm with H-minima labeling was used to segment the images at the optimal scale. The best feature subset for classification was filtered using the recursive feature elimination cross-validation approach, which extracts multiple spectral indices from the images.A random forest classifier combining superpixel segmentation and feature selection methods was proposed for the wetland classification. The model improves the classification accuracy of wetlands compared to three classical pixel-based machine learning classification methods. And the overall accuracy was 91.74% and the kappa coefficient was 0.9078, both of which were improved by about 4.53% and 0.0506, respectively, compared with the best-performing random forest classifier in pixel-oriented. The results showed that this method can effectively improve the classification accuracy of the Yellow River Delta wetlands compared with the previous studies.

Comparing Machine Learning Algorithms for Pixel/Object-Based Classifications of Semi-Arid Grassland in Northern China Using Multisource Medium Resolution Imageries

Knowledge of grassland classification in a timely and accurate manner is essential for grassland resource management and utilization. Although remote sensing imagery analysis technology is widely applied for land cover classification, few studies have systematically compared the performance of commonly used methods on semi-arid native grasslands in northern China. This renders the grassland classification work in this region devoid of applicable technical references. In this study, the central Xilingol (China) was selected as the study area, and the performances of four widely used machine learning algorithms for mapping semi-arid grassland under pixel-based and object-based classification methods were compared: random forest (RF), support vector machine (SVM), k-nearest neighbor (KNN), and naive Bayes (NB). The features were composed of the Landsat OLI multispectral data, spectral indices, Sentinel SAR C bands, topographic, position (coordinates), geometric, and grey-level co-occurrence matrix (GLCM) texture variables. The findings demonstrated that (1) the object-based methods depicted a more realistic land cover distribution and had greater accuracy than the pixel-based methods; (2) in the pixel-based classification, RF performed the best, with OA and Kappa values of 96.32% and 0.95, respectively. In object-based classification, RF and SVM presented no statistically different predictions, with OA and Kappa exceeding 97.5% and 0.97, respectively, and both performed significantly better than other algorithms. (3) In pixel-based classification, multispectral bands, spectral indices, and geographic features significantly distinguished grassland, whereas, in object-based classification, multispectral bands, spectral indices, elevation, and position features were more prominent. Despite the fact that Sentinel 1 SAR variables were chosen as an effective variable in object-based classification, they made no significant contribution to the grassland distinction.

Extraction of Broad-Leaved Tree Crown Based on UAV Visible Images and OBIA-RF Model: A Case Study for Chinese Olive Trees

Chinese olive trees (Canarium album L.) are broad-leaved species that are widely planted in China. Accurately obtaining tree crown information provides important data for evaluating Chinese olive tree growth status, water and fertilizer management, and yield estimation. To this end, this study first used unmanned aerial vehicle (UAV) images in the visible band as the source of remote sensing (RS) data. Second, based on spectral features of the image object, the vegetation index, shape, texture, and terrain features were introduced. Finally, the extraction effect of different feature dimensions was analyzed based on the random forest (RF) algorithm, and the performance of different classifiers was compared based on the features after dimensionality reduction. The results showed that the difference in feature dimensionality and importance was the main factor that led to a change in extraction accuracy. RF has the best extraction effect among the current mainstream machine learning (ML) algorithms. In comparison with the pixel-based (PB) classification method, the object-based image analysis (OBIA) method can extract features of each element of RS images, which has certain advantages. Therefore, the combination of OBIA and RF algorithms is a good solution for Chinese olive tree crown (COTC) extraction based on UAV visible band images.